Advertisement

Opioid Use Among Individuals With Spinal Cord Injury: Prevalence Estimates Based on State Prescription Drug Monitoring Program Data

Published:November 20, 2020DOI:https://doi.org/10.1016/j.apmr.2020.10.128

      Abstract

      Objective

      To identify the prevalence of opioid use in individuals with chronic spinal cord injury (SCI) living in South Carolina.

      Design

      Cohort study.

      Setting

      Data from 2 statewide population-based databases, an SCI Registry and the state prescription drug monitoring program, were linked and analyzed.

      Participants

      The study included individuals (N=503) with chronic (>1y) SCI who were injured between 2013 and 2014 in South Carolina and who survived at least 3 years postinjury.

      Interventions

      Not applicable.

      Main Outcome Measures

      Filled opioid prescriptions over a 2-year period (months 13-36 after injury). The main outcomes were total number of days with an opioid prescription over the 2-year period, length of coverage period [(final day of prescription coverage+the days supplied)–first day of prescription coverage], average daily morphine milligram equivalents (MME) over the coverage period, and concurrent days covered by an opioid and a prescription for benzodiazepines, sedatives, or hypnotics.

      Results

      A total of 53.5% of the cohort (269 individuals) filled at least 1 opioid prescription during their second or third year after SCI. In total, there were 3386 opioid fills during the 2-year study. On average, the total number of opioid prescription days was 293±367. The average coverage period was 389±290 days, and the average daily MME during the coverage period was 41±70 MME. Of those who filled an opioid prescription, 23% had high-risk fills (>50 MME), and 38% had concurrent prescriptions for benzodiazepines, sedatives, or hypnotics.

      Conclusions

      The prevalence of opioid use was high among individuals with chronic SCI, exceeding rates observed in the general population. Also concerning were the rates of high-risk fills, based on average daily MME and concurrent benzodiazepine, sedative, or hypnotic prescriptions. These findings, taken together with those of earlier studies, should be used by providers to assess and monitor opioid use, decrease concurrent high-risk medication use, and attenuate the risk of adverse outcomes.

      Keywords

      List of abbreviations:

      95% CI (95% confidence interval), CDC (Centers for Disease Control and Prevention), MME (morphine milligram equivalents), PDMP (prescription drug monitoring program), SCI (spinal cord injury), SCISRS (South Carolina SCI Surveillance and Registry System), SCRIPTS (South Carolina Reporting & Identification Prescription Tracking System)
      In light of the current opioid epidemic, which was declared a public health emergency by the United States in 2017,
      US Department of Health and Human Services
      Renewal of determination that a punblic health emergency exists, 2019.
      ,
      US Department of Health and Human Services
      What is the U.S. opioid epidemic? 2019.
      there has been increased focus on opioid use and risk of adverse outcomes (eg, opioid misuse, overdose, death) among those with spinal cord injury (SCI). Individuals with SCI commonly experience secondary health conditions, including pain, spasticity, and depression, which frequently result in polypharmacy and may involve concurrent treatment with high-risk prescription medications including opioids, benzodiazepines, and sedatives.
      • Jensen E.K.
      • Biering-Sorensen F.
      Medication before and after a spinal cord lesion.
      • Kitzman P.
      • Cecil D.
      • Kolpek J.H.
      The risks of polypharmacy following spinal cord injury.
      • Patel T.
      • Milligan J.
      • Lee J.
      Medication-related problems in individuals with spinal cord injury in a primary care-based clinic.
      • Hogholen H.
      • Storhaug A.
      • Kvernrod K.
      • Kostovski E.
      • Viktil K.K.
      • Mathiesen L.
      Use of medicines, adherence and attitudes to medicines among persons with chronic spinal cord injury.
      Concurrent use of these medications is concerning and generally cautioned against, because it increases the risk of overdose and death.
      • Dowell D.
      • Haegerich T.M.
      • Chou R.
      CDC guideline for prescribing opioids for chronic pain--United States, 2016.
      Pain, in particular, is a highly complex and prevalent condition often necessitating multimodal treatment with prescription opioid analgesics, psychotropic medications, off-label prescription medication use, and nonpharmacologic relief strategies.
      • Teasell R.W.
      • Mehta S.
      • Aubut J.A.
      • et al.
      A systematic review of pharmacologic treatments of pain after spinal cord injury.
      • Woller S.A.
      • Hook M.A.
      Opioid administration following spinal cord injury: implications for pain and locomotor recovery.
      • Hagen E.M.
      • Rekand T.
      Management of neuropathic pain associated with spinal cord injury.
      • Mehta S.
      • McIntyre A.
      • Janzen S.
      • Loh E.
      • Teasell R.
      Spinal Cord Injury Rehabilitation Evidence Team. Systematic review of pharmacologic treatments of pain after spinal cord injury: an update.
      • Hatch M.N.
      • Cushing T.R.
      • Carlson G.D.
      • Chang E.Y.
      Neuropathic pain and SCI: identification and treatment strategies in the 21st century.
      However, the management of chronic pain with prescription opioids is controversial and has gained increased attention in recent years.
      • New P.W.
      Severe chronic pain following spinal cord damage: a pragmatic perspective for prescribing opioids.
      ,
      • Bryce T.N.
      Opioids should not be prescribed for chronic pain after spinal cord injury.
      Although opioids are not recommended as a first-line therapy for pain after SCI (their use has been recommended only as a last resort), they are commonly prescribed and often at high dosages, resulting in an increased risk of adverse outcomes.
      • Kitzman P.
      • Cecil D.
      • Kolpek J.H.
      The risks of polypharmacy following spinal cord injury.
      ,
      • Hagen E.M.
      • Rekand T.
      Management of neuropathic pain associated with spinal cord injury.
      ,
      • Hatch M.N.
      • Cushing T.R.
      • Carlson G.D.
      • Chang E.Y.
      Neuropathic pain and SCI: identification and treatment strategies in the 21st century.
      Several studies have highlighted the increased risk of misuse,
      • Clark J.M.
      • Cao Y.
      • Krause J.S.
      Risk of pain medication misuse after spinal cord injury: the role of substance use, personality, and depression.
      ,
      • Krause J.S.
      • Clark J.M.
      • Saunders L.L.
      Pain medication misuse among participants with spinal cord injury.
      unintentional injuries (eg, falls and overdose),
      • Krause J.S.
      Factors associated with risk for subsequent injuries after the onset of traumatic spinal cord injury.
      ,
      • Krause J.S.
      • Cao Y.
      • DiPiro N.D.
      • Cuddy E.
      Personality, high-risk behaviors, and elevated risk of unintentional deaths related to drug poisoning among individuals with spinal cord injury.
      and mortality
      • Clark J.M.
      • Cao Y.
      • Krause J.S.
      Risk of pain medication misuse after spinal cord injury: the role of substance use, personality, and depression.
      • Krause J.S.
      • Clark J.M.
      • Saunders L.L.
      Pain medication misuse among participants with spinal cord injury.
      • Krause J.S.
      Factors associated with risk for subsequent injuries after the onset of traumatic spinal cord injury.
      • Krause J.S.
      • Cao Y.
      • DiPiro N.D.
      • Cuddy E.
      Personality, high-risk behaviors, and elevated risk of unintentional deaths related to drug poisoning among individuals with spinal cord injury.
      • Krause J.S.
      • Carter R.E.
      • Pickelsimer E.
      Behavioral risk factors of mortality after spinal cord injury.
      • Krause J.S.
      • Zhai Y.
      • Saunders L.L.
      • Carter R.E.
      Risk of mortality after spinal cord injury: an 8-year prospective study.
      • Krause J.S.
      • Cao Y.
      • Clark J.M.
      Pain intensity, interference, and medication use after spinal cord injury: relationship with risk of mortality after controlling for socioeconomic and other health factors.
      • Cao Y.
      • Clark J.M.
      • Krause J.S.
      Changes in psychotropic prescription medication use and their relationship with mortality among people with traumatic spinal cord injury.
      associated with prescription pain medication use after SCI. However, these studies focused on self-reported prescription medication use for pain, lacked detailed information regarding specific prescription use, such as type, dosage, and frequency, and are subject to recall bias, among other limitations. Few other studies have assessed the prevalence of opioid use and prescribing patterns. One study using data from the Veterans Affairs Spinal Cord Dysfunction Registry found upward of 70% of the male population with SCI were prescribed opioids for pain.
      • Carbone L.D.
      • Chin A.S.
      • Lee T.A.
      • et al.
      The association of opioid use with incident lower extremity fractures in spinal cord injury.
      Another study found that, compared to propensity-score matched controls, individuals with SCI were more likely to use opioids for longer durations and at higher morphine-equivalent dosages.
      • Hand B.N.
      • Krause J.S.
      • Simpson K.N.
      Dose and duration of opioid use in propensity score matched, privately insured opioid users with and without spinal cord injury.
      International surveys of providers, completed in 2018, shed light on the approaches to opioid prescribing from a clinical perspective. One survey found that nearly 75% of prescribing clinicians reported their patients took opioids for chronic pain; in developed nations, the percentage was higher (83%).
      • Stillman M.
      • Graves D.
      • New P.W.
      • Bryce T.
      • Alexander M.
      Survey on current treatments for pain after spinal cord damage.
      Another found that, of providers who believe opioids have a role in chronic pain treatment, nearly half think there should be no upper limit for the dosage and would continue prescribing high dosages if it were effective.
      • Wong T.K.
      • Alexander M.S.
      • New P.W.
      • Delgado A.D.
      • Bryce T.N.
      Pulse article: opioid prescription for pain after spinal cord damage (SCD), differences from recommended guidelines, and a proposed algorithm for the use of opioids for pain after SCD.
      To date, there are limited data characterizing prescription opioid use among those with chronic SCI and little is known regarding the prevalence of having high-risk concurrent prescriptions, including benzodiazepines, sedatives, or hypnotics. Taken together, these earlier findings highlight the need for enhanced studies of prescription opioid use, specifically using data from administrative sources, such as prescription drug monitoring programs (PDMPs), which may be used to assess prevalence, dosage, and prescribing patterns. PDMPs are statewide electronic databases that collect, monitor, and analyze data on substances dispensed within the state and are promising tools for improving prescribing patterns, informing health providers and authorities, and decreasing patient risk. To our knowledge, there has yet to be an analysis of data from a state PDMP to assess opioid use among individuals with SCI.
      Our purpose was to identify the prevalence of opioid prescription fills, prescribed dosages, and patterns of fills among a cohort of individuals with chronic SCI (>1y postinjury) by linking administrative records from 2 statewide databases.

      Methods

      Data sources

      South Carolina SCI Surveillance and Registry System

      The South Carolina SCI Surveillance and Registry System (SCISRS) is populated with the injury records from the Statewide Hospital Discharge Data Set, which was developed under the mandate for all nonfederal hospitals in the state to submit uniform billing (UB-04) data on all discharge records to the South Carolina Revenue & Fiscal Affairs Office, Health & Demographics Section. The population-based registry represents all incident cases of SCI in the civilian population of South Carolina. Individuals with SCI were identified using International Classification of Diseases–9th Revision–Clinical Modifications nature of injury codes 806.0-806.9 and 952.0-952.9 in accordance with the Centers for Disease Control and Prevention (CDC) central nervous system injury guidelines.
      • Butler J.
      • Langlois J.A.
      Central nervous system injury surveillance: annual data submission standards-2000.

      South Carolina Reporting & Identification Prescription Tracking System (SCRIPTS)

      The South Carolina PDMP, known as South Carolina Reporting & Identification Prescription Tracking System (SCRIPTS), captures prescriptions filled in or shipped into South Carolina for state residents. The state-mandated PDMP is useful for monitoring and evaluating prescription and dispensing data and is important for assessing provider prescribing patterns, prescription medication misuse and diversion, and opioid-related mortality. The South Carolina Prescription Monitoring Act

      South Carolina Code of Laws. South Carolina Prescription Monitoring Act. SECTION 44-53-1650 D (8). Confidentiality; persons to whom data may be released. Available at: https://scdhec.gov/prescription-monitoring-act. Accessed January 8, 2020.

      (Section 44-53-1650[D]) permits release of data to qualified personnel for “the purpose of bona fide research or education; however, data elements that would reasonably identify a specific recipient, prescriber, or dispenser must be deleted or redacted from such information prior to disclosure.”

      Participants

      Individuals who had a new SCI occurring between 2013 and 2014 who were injured and living in the state South Carolina and had survived their second and third year postinjury were eligible for inclusion. The SCISRS identified 510 individuals. Using updated death certificate data, 5 individuals were excluded due to death during the second and third year postinjury. Two other individuals were deleted due to small cell sizes when cross-tabulating data, leaving a final sample of 503 records from the SCISRS. Of these individuals, 269 were matched to SCRIPTS for having an opioid prescription filled during the study period, within the second and third years (months 13-36) after injury. The matching protocol was assessed and determined to be robust, that is, those from the SCISRS not found in SCRIPTS were unlikely to have had an opioid prescription during the study period.

      Prescription outcomes

      The resultant linked dataset was used to generate summary statistics to describe opioid use in the cohort of individuals with SCI. To minimize issues with confidentiality concerns related to small cell sizes when cross-tabulating matched data, several identifying variables were condensed, notably age, race, and SCI level. In the final linked dataset, each opioid prescription included the following variables: unique study ID, sex (men, women), age group (<40, 40-64, >64), race (white, non-white), SCI level (cervical, noncervical), months after SCI, days after SCI, prescription drug, days supplied, quantity, dosage, and morphine milligram equivalents (MME) for the prescription fill, provided by the PDMP. MME is a way of standardizing the strengths of different opioid prescription medications.
      Our primary outcome was receipt of at least 1 opioid prescription during the second or third year postinjury. The final dataset contained 13,196 prescription fills for all schedules II, III, and IV controlled substances reported in the PDMP for the 503 individuals during the study timeframe. Focusing only on the 269 individuals who had an opioid prescription filled in the second and third years after injury reduced it to 3386 opioid prescription fills.
      For those with filled prescriptions, we identified: (1) the total number of days supplied for opioid prescriptions (ie, the sum of prescribed opioid days); (2) the length of the coverage period (ie, the number of days an individual could have had an opioid in hand within the 2-year timeframe, calculated as: (final day of prescription coverage+the days supplied)–first day of prescription coverage; (3) the average daily MME over the prescription coverage period, which includes all opioid prescriptions that might be overlapping, calculated based on the sum of the total MME, divided by the coverage period; and (4) concurrent days covered by an opioid prescription and a prescription for benzodiazepines, sedatives, or hypnotics.
      It is important to note that a person might have had multiple overlapping opioid prescriptions in the same month, making the total number of opioid prescription days more than the actual number of days in the month. The prescription coverage period, on the other hand, considers all opioids the individual had access to during that timeframe, but it does not take into account gaps in coverage that may have occurred. The average daily MME data were further categorized into 3 levels: <50 MME, 50-90, and >90 MME, consistent with reporting by the Centers for Disease Control and Prevention to identify high average daily dose.
      Centers for Disease Control and Prevention
      Annual surveillance report of drug-related risks and outcomes—United States, 2019.

      Statistical analysis

      All analyses were completed with SAS version 9.4.a Descriptive statistics, n (%) or mean ± SD, are presented. Bivariate analyses of the relations between the personal characteristics and opioid outcomes are also presented. A logistic regression model was run to assess the relation between personal characteristics and having had a prescription for opioids during the study period (N=503). A multiple regression model (Proc Genmod with negative binomial distribution) was run to assess the relation between personal characteristics and average daily MME over the coverage period (n=269). Odds ratios and incidence rate ratios are reported with confidence intervals (CIs).

      Results

      Of the 503 individuals with SCI identified through the SCISRS, 53.5% had at least 1 opioid prescription listed in the SCRIPTS database in the second- or third-year postinjury. Participant characteristics of the total sample, those who received a prescription in year 2 or 3 postinjury, and those who did not are presented in table 1. The only significant finding was that women were more likely than expected to receive an opioid prescription.
      Table 1Participant characteristics, n (%)
      CharacteristicsTotal (N=503)Filled Opioid Rx (n=269)No Opioid Rx (n=234)
      Sex
       Men357 (71)177 (66)179 (77)
       Women146 (29)92 (34)
      Statistically significant at the P<.05 level.
      55 (24)
      Race
       White278 (55)152 (57)126 (54)
       Other225 (45)117 (43)108 (46)
      Age group
       <40148 (29)84 (31)64 (27)
       40-64210 (42)119 (44)91 (39)
       65+145 (29)66 (25)79 (34)
      Injury level
       Cervical303 (60)158 (59)145 (62)
       Noncervical200 (40)111 (41)89 (38)
      Statistically significant at the P<.05 level.
      The logistic regression showed that the odds ratios of having an opioid prescription were 1.92 (95% CI, 1.27-2.93; P=.002) for women, 1.83 (95% CI, 1.12-2.99; P=.004) for those younger than 40 years, and 1.94 (95% CI, 1.23-3.05; P=.02) for those between 40 and 64 years old, compared to those older than 65 (table 2).
      Table 2Logistic regression comparing personal characteristics of opioid users and nonusers (N=503)
      VariablesOdds Ratio95% CIP Value
      Sex (ref: men)
       Women1.921.272.93.002
      Significant.
      Race (ref: white)
       Other0.920.631.33.64
      Age group (ref: 65+)
       <401.831.122.99.004
      Significant.
       40-641.941.233.05.02
      Significant.
      Injury level (ref: noncervical)
       Cervical0.960.651.43.85
      Significant.
      By year postinjury, the estimated prevalence of prescription opioid use was 47% in year 2 and 40% in year 3 (table 3). Overall, and within each subpopulation, the prevalence of opioid use in the third-year postinjury was lower than in the second year postinjury. Again, the prevalence of opioid use was significantly higher among women compared to men. Differences by race, age, and injury level were not statistically significant.
      Table 3Estimated prevalence of prescription opioid use among persons with SCI during years 2 and 3 postinjury, n (%)
      Total 2-Year Cohort, nYear 2 Filled Opioid Rx n (%)Year 3 Filled Opioid Rx n (%)
      503234 (47)202 (40)
      Sex
       Men357152 (43)131 (37)
       Women14682 (56)
      Statistically significant at the P<.05 level.
      71 (49)
      Statistically significant at the P<.05 level.
      Race
       White278136 (49)119 (43)
       Other22598 (44)83 (37)
      Age group
       <4014874 (50)60 (41)
       40-64210102 (49)92 (44)
       65+14558 (40)50 (34)
      Injury level
       Cervical303135 (45)115 (38)
       Noncervical20099 (50)87 (44)
      Statistically significant at the P<.05 level.
      Among the 269 individuals with an opioid prescription listed in the PDMP during the 2-year study period, 3386 opioid prescriptions were filled; the mean number of fills per person was 12.59±13.71 (range=1-79). The average daily MME per prescription fill was 58.9, showing the amount of opioid dispensed to a patient. Comparatively, the average daily MME during the prescription coverage period (presented below) depicts potency across the total number of days covered by all fills.
      Over the 2-year study period (months 13-36 after injury), the total number of days covered by an opioid prescription per individual (ie, sum of prescribed opioid days) ranged from 1 to 1729 days (mean 292.86±367.13). The average prescription coverage period (index fill date to the last fill date, plus days supplied by the final fill) was 388.67±290.05 days (range=1-755d), and the average daily MME over the coverage period was 40.96±69.88 MME (range=0.17-750). Most individuals (77.3%) received <50 MME per day over the prescription coverage period. The distribution of persons by average daily MME over the prescription coverage period is presented in table 4. There were no significant differences between the groups.
      Table 4Distribution of individuals by average daily MME over the prescription coverage period (%)
      Opioid Rx n=269<50 MME/Day n=20850-89 MME/Day n=34≥90MME/Day n=27
      Sex
       Men65.967.766.7
       Women33.132.333.3
      Race
       White54.355.974.1
       Other45.744.125.9
      Age group
       <4030.844.118.5
       40-6443.841.251.9
       65+25.514.729.6
      Injury level
       Cervical62.047.148.2
       Noncervical38.052.951.9
      In addition to the 3386 opioid prescriptions, there were 1546 concurrent prescriptions for benzodiazepines, sedatives, and hypnotics. Thirty-eight percent of the individuals (n=103) had concurrent prescription fills for opioids and benzodiazepines, sedatives, or hypnotics over the 2-year period, and of those, one-quarter (n=26) had an overlap over 360 days.
      The bivariate analyses showed women had significantly longer prescription coverage periods as well as a higher number of days covered by concurrent prescriptions for opioids and benzodiazepines, sedatives, or hypnotics (table 5). White individuals had significantly more opioid prescription days, longer prescription coverage periods, higher average daily MME for the period, and higher number of concurrent days covered. There were no statistically significant differences among the age groups or injury levels.
      Table 5Bivariate relation between opioid users’ personal characteristics and outcomes (n=269)
      Total Days Covered by Opioid RxOpioid Rx Coverage Period
      Coverage period=(final day of prescription coverage+days supplied)–first day of prescription coverage.
      (d)
      Average Daily MME Over Coverage PeriodDays With Concurrent Opioid and BSH Rx
      Sex
       Men275.3±369.4363.9±290.437.1±48.460.7±156.0
       Women326.6±362.3437.1±284.7
      Statistically significant at the P<.05 level.
      48.6±99.2117.5±202.7
      Statistically significant at the P<.05 level.
      Race
       White339.6±397.0
      Statistically significant at the P<.05 level.
      419.4±280.5
      Statistically significant at the P<.05 level.
      49.6±87.2
      Statistically significant at the P<.05 level.
      114.6±208.0
      Statistically significant at .001 level.
       Other233.0±316.7348.7±298.529.8±34.136.0±106.2
      Age group
       <40241.4±377.8372.9±285.733.9±49.469.6±165.9
       40-64336.2±379.1402.2±287.441.6±63.482.4±186.5
       65+280.8±324.0384.5±303.548.8±98.089.6±167.0
      Injury level
       Cervical274.3±365.8370.6±297.535.4±64.571.3±166.0
       Noncervical319.3±369.0414.4±278.448.9±76.592.7±187.5
      NOTE. Values are given as mean ± SD.
      Abbreviation: BSH, benzodiazepines, sedatives, or hypnotics.
      Coverage period=(final day of prescription coverage+days supplied)–first day of prescription coverage.
      Statistically significant at the P<.05 level.
      Statistically significant at .001 level.
      Based on the results of the multiple regression analysis, individuals younger than 40 had significantly lower average daily MME compared to those 65 and older and white individuals, and those with a noncervical SCI had significantly higher average daily MME (table 6).
      Table 6Multiple regression of opioid users’ personal characteristics on average daily MME during period of coverage (n=269)
      VariablesIRR95% CIP Value
      Sex (ref: men)
       Women1.020.741.41.90
      Race (ref: white)
       Other0.610.460.82.0009
      Significant.
      Age group (ref: 65+)
       <400.640.420.96.03
      Significant.
       40-640.910.631.32.62
      Injury level (ref: noncervical)
       Cervical0.670.500.90.007
      Significant.
      Abbreviation: IRR, incidence rate ratio.
      Significant.

      Discussion

      This study is unique in the linkage of 2 statewide administrative databases, the SCISRS and SCRIPTS, to examine prescription opioid use among individuals with chronic SCI living in the state of South Carolina, finding high rates of opioid use among those in the second- and third-year postinjury. This is an important first step in exploring the complex patterns of prescription opioid use by persons with SCI. Considering the results of earlier studies and the limited data on opioid prescriptions from administrative sources and the clinical perspective surveys, the current findings may have significant implications for quality of life, pain management strategies, societal cost, and risk of adverse outcomes (overdose and death) related to opioid use after SCI.
      Overall, 53% of the individuals with SCI filled at least 1 prescription for an opioid during their second and third year postinjury between 2015 and 2017. Of the 503 individuals identified by the registry, 47% filled an opioid prescription in year 2 after injury, whereas 40% filled a prescription in year 3 after injury. Comparatively, in 2017, 17.4% of the United States population (14.8% of men and 19.9% of women) filled an opioid prescription. Consistent with national
      Centers for Disease Control and Prevention
      Annual surveillance report of drug-related risks and outcomes—United States, 2019.
      and South Carolina
      • Kelly S.
      • Thomson L.
      • Heidari K.
      • Sen N.
      Opioid prescriptions in South Carolina.
      data, we found that a larger percentage of women had filled opioid prescriptions. Based on the results from the logistic regression, compared to men, the odds of women having received an opioid prescription over the 2-year study period was 1.92. We also found that those younger than 65 had higher odds of having an opioid prescription.
      Looking at the data from a prescription standpoint and drawing crude comparisons, we found higher rates of opioid prescription fills, as well as higher MME per prescription fill, in those with SCI compared to the general South Carolina population. Based on 2017 data from the SCRIPTS, roughly 4.3 million opioids were dispensed to residents of South Carolina in 2017, and the overall opioid prescription fill rate was 86.3 prescriptions per 100 persons.
      • Kelly S.
      • Thomson L.
      • Heidari K.
      • Sen N.
      Opioid prescriptions in South Carolina.
      Comparatively, the overall annual rate in our SCI cohort was 336.6 per 100 persons. Those with SCI had higher average daily MME per prescription fill (58.9), as compared to the general South Carolina population, where the average daily MME per prescription was 45.9.
      • Kelly S.
      • Thomson L.
      • Heidari K.
      • Sen N.
      Opioid prescriptions in South Carolina.
      There were also observed differences in the rates of opioid prescription dosages of <50 MME, 50-90 MME, and >90 MME per 100 persons, with higher rates among those with SCI. South Carolina ranks 20th in the United States for high opioid dosage, based on the rate of opioid prescriptions resulting in daily dosage greater than 90 MME, with a rate of 6.2 prescriptions per 100 persons. In the SCI sample, we found the rate of >90 MME per day was 10 times higher, at 62.0 prescriptions per 100 persons.
      Further highlighting the risks of opioid use in this population, from the patient-centered stance, we found that during the prescription coverage period (the timeframe in which the individual had opioids available to them) 23% of the sample had daily MME exceeding 50 MME per day, which is associated with a twofold risk of opioid overdose, compared to 20 MME or less per day.
      Centers for Disease Control and Prevention
      Calculating total daily dose of opioids for safer dosage, 2018.
      The average daily MME over the coverage period, which was a conservative estimate, was nearly at the 50 MME threshold for women (48.6), whites (49.6), individuals older than 65 years (48.8), and those with noncervical injuries (48.9). The risks of high MMEs must be discussed with these individuals and their opioid use carefully assessed and monitored. The regression analyses substantiate the findings, showing that white individuals and those with noncervical SCI had significantly higher average daily MME for the coverage period, whereas those below age 40 had significantly lower average daily MME for the coverage period.
      The concurrent use of opioids and other high-risk medications (benzodiazepines, sedatives, hypnotics) is also a major concern that, while generally cautioned against, has not yet been thoroughly investigated in those with SCI.
      • Kitzman P.
      • Cecil D.
      • Kolpek J.H.
      The risks of polypharmacy following spinal cord injury.
      ,
      • Dowell D.
      • Haegerich T.M.
      • Chou R.
      CDC guideline for prescribing opioids for chronic pain--United States, 2016.
      Our findings suggest there was considerable overlap, with 38% of the sample having concurrent prescriptions, which can be unsafe and poses increased risk of overdose and death. This, taken with the findings on opioid use, should be used by providers to decrease concurrent high-risk medication use and attenuate the risk of adverse outcomes.

      Study limitations

      Several limitations should be acknowledged when interpreting the findings. The prescription data come from the PDMP and are limited by the nature of this data source. Notably missing are detailed individual demographic (age, race) and clinical data (eg, injury severity, residual functional impairments, diagnosis, clinical reasons for opioid use). Despite the need to collapse race, we retained the variable because we wanted to determine if and what type of racial disparities may exist around prescribed opioid use in this subset of the SCI population, because findings could lead to recommendations for future research that translates into better pain management through communications to prescribers and care givers. In addition, because the data came from PDMP records, they only capture what is prescribed and filled in the state, not what is obtained from border states or other sources. A small percentage of individuals in the SCI registry (6%) live in border counties and may use providers and pharmacies outside the state; therefore, our data may underestimate use in this cohort. Importantly, we cannot account for actual use and must acknowledge that some individuals may misuse the prescriptions—some may not actually take the prescriptions as prescribed, some may be using more than what they were prescribed (eg, taking or borrowing from others), and others may sell or divert the medications. We also cannot account for discontinuation of use due to negative side effects or simply because the individuals did not like them or no longer needed for them. There are a few limitations specifically related to the prescription coverage period that may affect other outcomes presented. The period was limited by the 2-year study duration and may not have included prescriptions the individual had on hand from the days or month prior to the study start date. In addition, the coverage period did not account for gaps in coverage and therefore may have overestimated the duration of opioid availability. Irregular refills and/or a high number of gap days would lead to a lower calculated MME over the coverage period, which would not accurately represent the peak MME usage and may underestimate the risk of high usage and adverse outcomes. Last, we defined chronic SCI as having occurred at a minimum of 1 year earlier, and we do not have information on whether some participants may have had a complete or relatively complete recovery. Therefore, the findings apply to all individuals who were initially treated for acute SCI who are now a minimum of 1 year postinjury.
      Several strengths of the data are worth acknowledging as well. Because the data came from a PDMP, we have more accurate opioid prescription details compared to self-report, and the risk of recall bias has been eliminated. In addition, the population-based data are superior in comparison to clinical cohorts, because it captures everyone within a geographic region. Clinical cohorts are generally limited to a specific condition and institution and cannot be used to develop prevalence estimates.

      Future research

      Building on our findings, continued study of pain and the drivers of prescription opioid use are warranted. Future research should assess trends in opioid use among those with SCI, specifically whether there have been reductions in opioid prescriptions given the increased attention from the opioid epidemic, and, if so, the effect of changes in prescribing patterns on pain intensity, pain interference, and other secondary health conditions. This research should determine the characteristics of individuals for whom changes in prescription opioid use result in favorable versus unfavorable outcomes. Greater research is necessary to understand the risk-to-benefit ratio and the effectiveness of opioid use compared to other pain management modalities.

      Conclusions

      We found a high prevalence of opioid prescription fills, as well as high dosages among individuals living with chronic SCI in South Carolina, higher than observed in the general population. The rates of high-risk fills, based on MME and concurrent benzodiazepine, sedative, or hypnotic use, place these individuals at increased risk for adverse outcomes. Taken together with earlier findings, these data may be used by health care providers and researchers to assess and monitor opioid use better, decrease concurrent high-risk medication use, and attenuate the risks of adverse outcomes, including misuse, overdose, and death.

      Supplier

      • a.
        SAS version 9.4; SAS Institute.

      References

        • US Department of Health and Human Services
        Renewal of determination that a punblic health emergency exists, 2019.
        (Available at:) (Accessed January 8, 2020)
        • US Department of Health and Human Services
        What is the U.S. opioid epidemic? 2019.
        (Available at:) (Accessed January 8, 2020)
        • Jensen E.K.
        • Biering-Sorensen F.
        Medication before and after a spinal cord lesion.
        Spinal Cord. 2014; 52: 358-363
        • Kitzman P.
        • Cecil D.
        • Kolpek J.H.
        The risks of polypharmacy following spinal cord injury.
        J Spinal Cord Med. 2017; 40: 147-153
        • Patel T.
        • Milligan J.
        • Lee J.
        Medication-related problems in individuals with spinal cord injury in a primary care-based clinic.
        J Spinal Cord Med. 2017; 40: 54-61
        • Hogholen H.
        • Storhaug A.
        • Kvernrod K.
        • Kostovski E.
        • Viktil K.K.
        • Mathiesen L.
        Use of medicines, adherence and attitudes to medicines among persons with chronic spinal cord injury.
        Spinal Cord. 2018; 56: 35-40
        • Dowell D.
        • Haegerich T.M.
        • Chou R.
        CDC guideline for prescribing opioids for chronic pain--United States, 2016.
        JAMA. 2016; 315: 1624-1645
        • Teasell R.W.
        • Mehta S.
        • Aubut J.A.
        • et al.
        A systematic review of pharmacologic treatments of pain after spinal cord injury.
        Arch Phys Med Rehabil. 2010; 91: 816-831
        • Woller S.A.
        • Hook M.A.
        Opioid administration following spinal cord injury: implications for pain and locomotor recovery.
        Exper Neurol. 2013; 247: 328-341
        • Hagen E.M.
        • Rekand T.
        Management of neuropathic pain associated with spinal cord injury.
        Pain Ther. 2015; 4: 51-65
        • Mehta S.
        • McIntyre A.
        • Janzen S.
        • Loh E.
        • Teasell R.
        Spinal Cord Injury Rehabilitation Evidence Team. Systematic review of pharmacologic treatments of pain after spinal cord injury: an update.
        Arch Phys Med Rehabil. 2016; 97: 1381-1391. e1381
        • Hatch M.N.
        • Cushing T.R.
        • Carlson G.D.
        • Chang E.Y.
        Neuropathic pain and SCI: identification and treatment strategies in the 21st century.
        J Neurol Sci. 2018; 384: 75-83
        • New P.W.
        Severe chronic pain following spinal cord damage: a pragmatic perspective for prescribing opioids.
        Spinal Cord Ser Cases. 2018; 4: 65
        • Bryce T.N.
        Opioids should not be prescribed for chronic pain after spinal cord injury.
        Spinal Cord Ser Cases. 2018; 4: 66
        • Clark J.M.
        • Cao Y.
        • Krause J.S.
        Risk of pain medication misuse after spinal cord injury: the role of substance use, personality, and depression.
        J Pain. 2017; 18: 166-177
        • Krause J.S.
        • Clark J.M.
        • Saunders L.L.
        Pain medication misuse among participants with spinal cord injury.
        Spinal Cord. 2015; 53: 630-635
        • Krause J.S.
        Factors associated with risk for subsequent injuries after the onset of traumatic spinal cord injury.
        Arch Phys Med Rehabil. 2004; 85: 1503-1508
        • Krause J.S.
        • Cao Y.
        • DiPiro N.D.
        • Cuddy E.
        Personality, high-risk behaviors, and elevated risk of unintentional deaths related to drug poisoning among individuals with spinal cord injury.
        Arch Phys Med Rehabil. 2018; 99: 1941-1948
        • Krause J.S.
        • Carter R.E.
        • Pickelsimer E.
        Behavioral risk factors of mortality after spinal cord injury.
        Arch Phys Med Rehabil. 2009; 90: 95-101
        • Krause J.S.
        • Zhai Y.
        • Saunders L.L.
        • Carter R.E.
        Risk of mortality after spinal cord injury: an 8-year prospective study.
        Arch Phys Med Rehabil. 2009; 90: 1708-1715
        • Krause J.S.
        • Cao Y.
        • Clark J.M.
        Pain intensity, interference, and medication use after spinal cord injury: relationship with risk of mortality after controlling for socioeconomic and other health factors.
        Arch Phys Med Rehabil. 2017; 98: 2464-2470
        • Cao Y.
        • Clark J.M.
        • Krause J.S.
        Changes in psychotropic prescription medication use and their relationship with mortality among people with traumatic spinal cord injury.
        Spinal Cord. 2018; 56: 680-686
        • Carbone L.D.
        • Chin A.S.
        • Lee T.A.
        • et al.
        The association of opioid use with incident lower extremity fractures in spinal cord injury.
        J Spinal Cord Med. 2013; 36: 91-96
        • Hand B.N.
        • Krause J.S.
        • Simpson K.N.
        Dose and duration of opioid use in propensity score matched, privately insured opioid users with and without spinal cord injury.
        Arch Phys Med Rehabil. 2018; 99: 855-861
        • Stillman M.
        • Graves D.
        • New P.W.
        • Bryce T.
        • Alexander M.
        Survey on current treatments for pain after spinal cord damage.
        Spinal Cord Ser Cases. 2019; 5: 14
        • Wong T.K.
        • Alexander M.S.
        • New P.W.
        • Delgado A.D.
        • Bryce T.N.
        Pulse article: opioid prescription for pain after spinal cord damage (SCD), differences from recommended guidelines, and a proposed algorithm for the use of opioids for pain after SCD.
        Spinal Cord Ser Cases. 2019; 5: 39
        • Butler J.
        • Langlois J.A.
        Central nervous system injury surveillance: annual data submission standards-2000.
        US Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Injury Prevention and Control, Atlanta, GA2001
      1. South Carolina Code of Laws. South Carolina Prescription Monitoring Act. SECTION 44-53-1650 D (8). Confidentiality; persons to whom data may be released. Available at: https://scdhec.gov/prescription-monitoring-act. Accessed January 8, 2020.

        • Centers for Disease Control and Prevention
        Annual surveillance report of drug-related risks and outcomes—United States, 2019.
        (Available at:) (Accessed January 7, 2020)
        • Kelly S.
        • Thomson L.
        • Heidari K.
        • Sen N.
        Opioid prescriptions in South Carolina.
        South Carolina Department of Health and Environmental Control Bureau of Drug Control, 2018 (Available at https://scdhec.gov/sites/default/files/media/document/Opioid_Prescription_in_South_Carolina_Oct-2018.pdf. Accessed January 8, 2020)
        • Centers for Disease Control and Prevention
        Calculating total daily dose of opioids for safer dosage, 2018.
        (Available at:) (Accessed January 8, 2020)